¹ ‘Language as a natural object’, Jacobsen Lecture, May 23rd, 1994, part 1 of ‘Language and Nature’ op. cit. In this Mind version, the passage I have quoted occurs at page 24. See also pages 13, 29 (where Chomsky quotes Davidson to the same effect), 41, 48–49. From now onwards, square bracketed references are to this version.Google Scholar

² Languages do not have to change. Unless you count as change the regularization of spelling or the simple taking in of new words, Old Norse (Icelandic) has barely changed in two millennia. But languages are by their nature changeable.

³ To conceive that one variant could be a variant of two distinct languages is to conceive that the individuation and differentiation of L₁ and L₂ need not depend upon some locally evaluable discontinuity among variants of languages L₁ and L₂. Compare the differentiation of the determinables red and yellow. For determinables and their determinations or determinates, see Johnson, W. E., Logic volume 1, p. 174 follg.Google ScholarFor the claim that colours and their hues are particular, not general things and the explanation of how the predication of colours is possible and the predication of colours of colours is possible, see my ‘Verbs and Adverbs, and some other Modes of Grammatical Combination’, Aristotelian Society Proceedings, vol LXXXVI 1985-1986, section X ad fin.3.Google Scholar

⁴ The common sense view of languages may benefit from a perfectly general point which is consistent with the definiteness of identity: that one can make it clear which object one means without settling there and then every identity and/or difference question about it. (Or so I claim in ‘Singling out an object determinately’Google Scholar in Pettitt, and McDowell, (Editors) Language, Thought and Context Oxford 1985.) Questions of sameness and difference are rarely if ever to be settled by routine. There is no limit to the empirical information that may be needed for their determination, case by case. But it is surely not foolish to wonder here whether, in virtue of old discoveries of general laws in historical linguistics concerning vowel shifts, consonant changes, etc, taken perhaps in conjunction with new discoveries such as Bickerton's about the ways in which pidgin languages can settle down and enter the condition of stable (‘creole’) languages, the prospects may be better now for the understanding of questions of language identity than they can appear for the corresponding questions about bicycle-identity or road-identity or sewing machine-identity. Surely the prospects are better, moreover, for the understanding of the individuation and differentiation of public languages than they are for questions of idiolect identity. Do I now speak the same idiolect, I ask myself, as I spoke in 1966, at the time when I was perfectly innocent of all sorts of problems of philosophical communication which I am now better aware of (and have made countless if insufficient allowances for)? When one tries to answer that sort of question about idiolects and seeks to discover what on earth is meant to be at issue here (unless it is the question of what makes a carbon copy), one will soon reach for the thought that talk of idiolects needs to be abandoned in favour of talk about different ways of speaking the shared common language we all speak and do ourselves use only one fraction of. Indeed if anything cries out for ontological condemnation, one finds oneself saying, it is the particular idiolect, not the particular language.Google Scholar

⁵ This is to say that the common sense view is not only anti-essentialist, and historicist. It is also anti-reductive. Better, it will insist upon honesty in question of reduction. There are well considered rules about what it takes to achieve reduction. (See Tar ski, Woodger, Hempel, Nagel.) These rules set standards that are hard to satisfy—as, in the long run and in the nobility of their failure, British phenomenalists so painfully discovered. Why should one suppose that, if a serious attempt were made to reduce talk of languages to talk of idiolects (in some sense of ‘idiolect,’ to be determined by the individualist, along with the specification of the biconditionals that were the recipe for the reduction), then the outcome would be any more favourable to the reductive claim than it was in the case of phenomenalism?

⁶ Except perhaps in so far as this aim helps to give grammatical proposals a naturalness that speakers themselves can recognize (e.g. as in accordance with their own conscious efforts to divine the composition and parsing of obscure or ill-made utterances).Google Scholar

⁷ The comparison with justice is in one way far fetched. The aim of acting justly in such and such present situation is a different sort of aim (I know and do not need to be reminded) from the aim of going on record here now with a correct statement to the effect that such and such. The normativity that is involved in the two cases is different. But the point of the comparison in no way depends on these things being the same. It resides in the utter implausibility of any general presumption to the effect that anything and everything that depends for its identifiability on X must be reducible to X.Google Scholar

⁸ See John, Foster, ‘Meaning and Truth Theory’ in Gareth, Evans and John, McDowell (Editors), Truth and Essays in Semantics (Oxford, 1976):Google Scholar and see Crispin, Wright's critique of Foster in Theories of Meaning and Speaker's Knowledge’, in his collection, Realism, Meaning and Truth (Oxford: Blackwell, 1986, 1993).Google Scholar

⁹ As Plato, Carnap, Montague and countless others have envisaged too.

¹⁰ Or should one say ‘only with respect to things that they believe are red’? It is no accident, I think, that we are uncomfortable here.

¹¹ Let me record that Wright himself is not himself an enthusiast for the story. This is not because he sides with Foster, but because of general doubts that he has entertained about the current preconceptions of philosophical semantics. Like Foster, Wright doubts whether it even makes sense to ascribe to people intentions that they cannot articulate, and he doubts that the appeal to ‘implicit knowledge’ can really explain speakers' capacities. He asks whether, once Foster's conception is rejected, there is anything for anyone to do here except to describe the causal structure of the dispositions that correspond to the truth-conditional theorems that record what sentences mean. My comments will be readily reconstructible on the idea that such a description could dispense with all reference to a particular public language.

¹² See Sameness and Substance, (Blackwell: Oxford 1980),Google Scholar (second edition, Cambridge 1998), Chapter Three. An important further contribution to these subjects which will help us to live with and expect the appearance (though only the appearance) of indeterminacy is Timothy, Williamson; Identity and Discrimination (Blackwell, 1990) pp.137–41.Google Scholar

¹³ Which I should claim, however mutely and ingloriously, to have anticipated, see Sameness and Substance, preface.Google Scholar

¹⁴ The relativity is relativity to human will, or to what Chomsky calls ‘internal reference frames’ or whatever, not of course relativity to the interests of a serious scientist who is in search of projectible predicates that figure in generalizations that are fruitful for prediction or control.

¹⁵ This is not to say that the common sense conception will let pass the question of what the recipe will be for dispensing with the sentence ‘Chinese is the language of Beijing and Hong Kong but not Melbourne”. Surely we cannot allow this or similar facts simply to disappear without trace from the objective world-view.

¹⁶ Noam, Chomsky in Philosophical Topics 20.1 (1992), page 226. If I had known earlier of this exchange between Chomsky and Putnam, it would have expedited my opposition to Chomsky's conception. In the event, there is a pleasing complementarity and/or convergence between Putnam's and my reactions to his position.Google Scholar

¹⁷ The claim condenses a great deal of Aristotelian insight. But see e.g. Metaphysics Z.4–5.Google Scholar

¹⁸ When that standard is satisfied, moreover, many of the other supposed requirements can fall away.

¹⁹ For the examples, see Amis, K., ‘Getting it Wrong’, in Michaels, L. and Ricks, C. B. (edd.), The State of the Language, (University of California Press, 1980).Google Scholar

²⁰ These categories deserve to be distinguished. I take the predicate ‘Martian’ to signify, in its application to terrestrial beings, the incapacity to enter into sensibilities (with correlative meanings) that can only be culturally and/or interactively imparted.